AgCast by Purdue Commercial Agriculture
Jul 23, 2026 · 43 min · 11 segments
Farm equipment markets are constantly changing, but what can auction data tell us about where agriculture is headed? In this episode of the Purdue Commercial AgCast, hosts Chad Fiechter and Todd…
Andy CampbellGuest
Todd KuetheHost
Chad FiechterHost
The origins here is that, Andy, you allowed us the opportunity to use tractor zoom data to at least get a start.
And so anyways, maybe I can hit some high points, right? So what we did is we created this model where we were estimating sort of like the characteristics of the tractor.
How does it relate to the auction prices? And like just quick off the top, it was, you know, as we get into those hour buckets, the sort of usage categories of tractors, which we defined fairly coarsely, but it was the penalty or the price ratio.
decrease that you would see for an additional hour, essentially kind of worked in half from maybe like a hundred, in the data we're using, we said it was $120 for less than 500 hours and then $60 for between, I think it was 500 and a thousand.
I mean, it's a similar shape of curve that I see on spares, that I see on combines, tractors, any kind of hour type of machine.
That's typically what I see is that almost exponentially decreasing rate of depreciation.
So you ran a hedonic price, which is basically, the idea is you're buying a bundle of characteristics.
You're buying a set of bedrooms, set of bathrooms, kitchen, attached garage, whatever, right? And each of those things you're buying in that big price.
You're basically decomposing through regression, right, into those pieces, right? That's what you're doing.
So my question to Andy, like, Was that sort of way we think of things as an economist? Was that surprising to you, expected to you?
So, yeah, throwing in usage in there and then also age, any of these other quantifiable things and then the non quantifiable central like chat, I think you pulled out like, OK, we need to have brands as potential inputs and control variables yes so that was one and then also the types of auction too i knew that that was important from looking at the data in the past i've done simple sorting and i could see that there was a difference but to be able to prove that it was a difference in this model i thought was interesting but yeah not brand new
The origins here is that, Andy, you allowed us the opportunity to use tractor zoom data to at least get a start.
And so anyways, maybe I can hit some high points, right? So what we did is we created this model where we were estimating sort of like the characteristics of the tractor.
How does it relate to the auction prices? And like just quick off the top, it was, you know, as we get into those hour buckets, the sort of usage categories of tractors, which we defined fairly coarsely, but it was the penalty or the price ratio.
decrease that you would see for an additional hour, essentially kind of worked in half from maybe like a hundred, in the data we're using, we said it was $120 for less than 500 hours and then $60 for between, I think it was 500 and a thousand.
I mean, it's a similar shape of curve that I see on spares, that I see on combines, tractors, any kind of hour type of machine.
That's typically what I see is that almost exponentially decreasing rate of depreciation.
So you ran a hedonic price, which is basically, the idea is you're buying a bundle of characteristics.
You're buying a set of bedrooms, set of bathrooms, kitchen, attached garage, whatever, right? And each of those things you're buying in that big price.
You're basically decomposing through regression, right, into those pieces, right? That's what you're doing.
So my question to Andy, like, Was that sort of way we think of things as an economist? Was that surprising to you, expected to you?
So, yeah, throwing in usage in there and then also age, any of these other quantifiable things and then the non quantifiable central like chat, I think you pulled out like, OK, we need to have brands as potential inputs and control variables yes so that was one and then also the types of auction too i knew that that was important from looking at the data in the past i've done simple sorting and i could see that there was a difference but to be able to prove that it was a difference in this model i thought was interesting but yeah not brand new
The rest of this transcript — segmented and speaker-labeled, so you land on the exact moment something was said
Search every transcript — by keyword, by phrase, or by meaning, across every show Radar indexes
Trends — what is surging across podcasts, measured against its own baseline
Alerts — when a name you follow appears in a newly indexed episode
No account is needed to search Radar.